Podcast Atlas
concept Updated 2026-08-25 Topics: History, Politics

Coalition Self-Preservation Failure / 联盟自保失灵

Coalition self-preservation failure / 联盟自保失灵 is the alliance pattern 《资治通鉴·周纪》45｜以史为鉴，君王到底该不该杀人成性？ draws from the 318 BCE five-state attack on 秦国. 楚国, 赵国, 魏国, 韩国, and 燕国 reach 函谷关 with 楚怀王 as alliance leader, while 齐国 promises support but has 孟尝君 / 田文 make the Qi army move slowly.

The coalition has a clear common enemy, but its members do not convert that shared threat into shared risk. Each state wants the others to absorb the first attack and preserve its own strength. Qin exploits the hesitation by sending unusual troops to cut Chu’s grain route; when Chu withdraws, the rest of the army collapses with it.

《资治通鉴·周纪》46丨骂你是个“二百五”的历史来源 extends the concept into aftermath. After the coalition withdraws, 秦国 pursues 韩国 and 魏国 under 樗里子, defeats Han at 修鱼之战, and continues to 浊泽. The failure is therefore not only that the coalition fails to take Qin; it also leaves front-line members exposed to Qin’s punishment.

The concept extends 共同敌人联盟 and 合纵 by naming their failure mode. Fear of Qin can create direction, but a coalition still fails if every member treats the alliance as insurance paid for by someone else’s losses.

Hanji 953 adds the Suanzao version of the pattern. The anti-董卓 armies gather in large numbers but mostly drink and wait, leaving 曹操 to test the line alone at Bianshui. Even after Cao Cao explains a concrete pass-control plan, the coalition refuses the coordinated cost, and the Suanzao camp disperses under grain pressure.

Hanji 950 supplies the diagnostic prelude to that failure. The 反董卓联盟 has enough moral pressure to form and enough prestige to select 袁绍 as盟主, but 郑泰 predicts that the commanders’ separate status, self-appointment, weak hierarchy, and private calculations will stop them from sustaining unified action.

《资治通鉴·周纪》50丨张仪是怎样把楚怀王忽悠瘸的 is not a battlefield self-preservation case, but it extends the same alliance-fragility problem. 张仪 makes 楚怀王 prefer a promised private gain in 商於 over maintaining the 齐国 tie, so Qin breaks the shared front before the coalition has to decide who bears the first military cost.

《资治通鉴·周纪》94丨杀神白起，长平之战中的“草包”名将？（2） adds a non-army version at Changping. The eastern states do not assemble a rescue coalition after Zhao is encircled, and Qi refuses the grain request that might keep Zhao’s army alive. The failure is no longer only “each ally waits for another to fight first”; it can also be “each state waits while a shield state starves.”

《资治通鉴·周纪》95丨长平之战后秦国为什么没有灭了赵国（2） adds the pre- and post-defeat mentality behind that failure. 虞卿 tries to make a Zhao-Qi land transfer create 合纵 pressure, but the episode itself doubts 齐国’s willingness. It then shows 魏安釐王’s court treating Qin’s attack on 赵国 as potentially good for 魏国, until 孔斌 warns that Qin’s appetite will not stop with Zhao.

《资治通鉴·周纪》96丨毛遂自荐（2） adds a partial-rescue version. 魏安釐王 does send 晋鄙 with 100,000 troops toward 邯郸, but when 秦昭襄王 warns that Wei will be next after Zhao, the Wei army stops at 邺城. This is not simple non-participation; it is a coalition member moving enough to signal help, then preserving itself before decisive contact.

《资治通鉴·周纪》96丨毛遂自荐（3） adds the persuasion response to that failure. Because Wei has already shifted from rescue to status-submission bargaining through 新垣衍, 鲁仲连 tries to make the preservation strategy look self-defeating: honoring Qin as emperor may avoid one battle, but it exposes Wei’s court and ministers to deeper Qin control.

《资治通鉴·周纪》96丨毛遂自荐（6） adds the coercive workaround. Normal Wei policy remains self-preserving, so 信陵君 uses 窃符救赵 to force Wei’s military capacity into the coalition. The result succeeds against Qin, but it is not healthy alliance governance; it relies on stolen credentials and the killing of 晋鄙.

Qinji 130-3 adds a restored-state rebellion version. 项梁 asks Qi and Zhao to join a concentrated attack on 章邯, but 田荣 conditions Qi participation on Chu killing 田假 and Zhao killing 田角 plus 田间. Chu and Zhao refuse to kill political refugees under their protection, so the coalition preserves local moral and status commitments while losing operational unity against Qin.

Qinji 134 adds the completed 巨鹿之战 crowding-and-watching case. 陈馀 preserves his outside Zhao force until Zhang Yan and Chen Ze force a limited sortie that is wiped out, while other lords build camps near Julu but watch from their walls as 项羽 attacks. The coalition only joins after Chu has made Qin break.

This case is not a total failure, because Xiang Yu’s victory eventually activates the lords. But their participation comes after someone else has paid the decisive risk, so Julu turns coalition self-preservation into a post-victory submission pattern.

Key Claims

  • Hanji 950 adds that coalition self-preservation can be visible at formation, before battlefield inaction, when leaders have no credible mechanism for hierarchy, cost-sharing, or binding sacrifice.
  • Hanji 953 adds that a coalition can possess a correct operational plan and still fail if no member accepts first risk and shared logistics.
  • A common enemy is not enough to produce operational unity.
  • Shared slogans can hide unequal willingness to spend troops, food, and political risk.
  • Slow-walking support can be formally consistent with alliance language while functionally weakening it.
  • The first member to suffer a supply shock or morale break can trigger cascading withdrawal.
  • Anti-Qin hezong needs enforcement and aligned incentives, not only a named leader.
  • Coalition failure can shift from collective retreat into concentrated retaliation against the most exposed members.
  • Zhouji 50 adds a pre-battle analogue: a coalition can be severed by private promised gain before battlefield self-preservation even appears.
  • Zhouji 94 part 2 adds a passive-relief analogue: states can preserve themselves by withholding food or intervention until a buffer state’s army collapses.
  • Zhouji 95 part 2 adds a neighbor-opportunism analogue: a state can interpret another weak state’s disaster as benefit even when the stronger state’s expansion logic threatens both.
  • Zhouji 96 part 2 adds a stopped-relief analogue: a state may mobilize aid but halt before the decisive battlefield once retaliation risk becomes explicit.
  • Zhouji 96 part 3 adds that coalition self-preservation can be challenged by reframing appeasement as personal and institutional exposure rather than safety.
  • Zhouji 96 part 6 adds that a stalled coalition can be reactivated by irregular coercion, but that fix carries institutional and moral costs.
  • Qinji 130-3 adds that a coalition can fail when one member demands settlement of an internal succession feud before contributing troops against the common enemy.
  • Qinji 134 adds that a coalition can withhold decisive help until one member has already broken the enemy, then submit to the winner’s command.

Connections